Infant outcome after prenatal exposure to antiepileptic compounds is of considerable concern to women with epilepsy. Under most circumstances it would be detrimental to the woman's health, and perhaps the health of the fetus, to have the antiepileptic medication withdrawn during pregnancy. At the same time, there is risk to the fetus exposed to these compounds and minimizing teratogenicity is crucial. The problem is compounded if the epileptic woman has multiple seizure types or seizures refractory to monotherapy. Under these circumstances, polytherapy is required to control seizures and the rate of teratogenicity is typically two- to three-fold higher with multiple drugs. In rodent systems, growth and other physical anomalies apparent in the embryo or at birth, resulting from carbamazepine exposure during gestation, have been linked to reactive metabolite exposure. However, no studies have sought to identify the cause of learning and behavioral deficits which become apparent after birth in humans. Although carbamazepine (CBZ) forms at least one reactive arene oxide metabolite, the 10, 11-epoxide, few studies have been conducted to link any form of CBZ-induced teratogenicity to reactive metabolite exposure, though the likelihood of this mechanism has been recognized. Our laboratory has sought to test this hypothesis indirectly by utilizing a second drug, stiripentol (STP), which in polytherapy with CBZ reduces the formation of carbamazepine's 10, 11-epoxide while maintaining the level of the parent compound. This approach, while yielding important information with regards to a potentially safer clinical prescription, does not directly address the risk of the 10, 11-epoxide alone. The goal of the proposed research is to directly test the teratogenicity (in the form of growth problems at birth and developmental delays after birth) of carbamazepine's 10, 11-epoxide. This will be accomplished by administration of synthesized CBZ 10, 11-epoxide to pregnant monkeys. Growth and developmental comparisons will then be made among 2 groups of prenatally exposed infants emanating from this proposal l) synthesized CBZ 10, 11-epoxide and 2) vehicle controls. These groups can then be compared-to previously tested infants who here prenatally exposed to either l) GBZ (Tegretol (R)) monotherapy, 2) CBZ (Tegretol (R)) plus STP, and 3) STP monotherapy.